Introduction
Juven’s mission is to empower the membership ecosystem of organizations and to connect them to create new value. As one of the key impact areas for organizations is event management, Juven has taken up the challenge of providing an end-to-end solution to bring event management- and thereby overall organizational engagement- to the next level.
In this series of article, Juven will shed light on AI-powered facial recognition technology, the current state of its use in the event management market, our product and advantages, and our development roadmap.
Facial Recognition Check-in
Facial recognition technology is no longer confined to the realms of science fiction, nor is it only a mere industry showpiece anymore. It’s here now — from airports to hotels, to workplaces, facial recognition technology is being used to check in travelers, guests and employees in a smooth and efficient way.
Obviously, there is a lot of interest for this technology in the event management as it has huge commercial potential. It’s no wonder that cutting-edge tech startups that offer development in AI vision, like Megvii Face++ and SenseTime have raised hundreds of millions of dollars. But as of yet, there is no handy consumer grade product despite its availability in the commercials sector. This is especially strange considering how this is not even the very cutting-edge of technology anymore.
There are plenty of reasons to use facial recognition for check-ins. For one, it is really fast. One event agency described it as being 50 % faster than traditional check-in. It’s extremely secure, as advances in facial recognition technology using AI-machine learning means that they are very good at recognizing faces and are very unlikely to make mistakes checking in guests. Finally, the biggest reason to use facial recognition technology to check your guests in is the WOW-factor.
There are other reasons as well, and if you want to know more details, check out this white paper about the use of facial recognition technology in events by EventsMB.
Given these promising reasons and potential for deployment of this technology in the event space, many event management companies have indicated their interest in developing and deploying it.
Two technological developments have been crucial to the development of AI-vision for check-in purposes — The Internet of Things (IoT) and Cloud Computing.
The Internet of Things refers to the network of physical objects embedded with processing power and internet connectivity. Cisco System estimates that the number of smart hardware to people ratio had gone from 0.08 in 2003 to 1.84 in 2010. That number has only increased. And as digital miniaturization continues, these devices are getting smarter and smarter.
With cloud computing, the data received from the IoT can be collected and processed centrally and remotely in the cloud. Allowing deep learning AI algorithms access to reams and reams of data is necessary to train them to do things. In the case of AI-vision, it is to train them to recognize objects, such as chairs, and distinguish faces in images.
By putting the two together, one can have a cloud-computing AI for all kinds of services, though in this post it is AI-Vision which interests us.
Cloud-based AI-Vision: Let’s demystify it
Step 1: Write an application that can take photos and store them.
Step 2: The images are sent to and stored onto the cloud platform with AI capacities before the image processing, aka facial recognition, takes place.e.g. AWS or Google Cloud
Step 3: The user (usually the system administrator or onsite helper) sends the face image(s) captured on the app to be compared to the cloud platform, either one at a time or in batch, requesting a certain facial recognition task to be run.
Step 4: The cloud platform (not the app) will run the algorithms to determine the result of the request. It could be a yes/no response or choosing one image out of many. The result is sent back to the app.
Step 5: The app shows the result to the system administrator.
The key points to note are that the AI processing takes place on cloud and has nothing to do with the app that is at use except for the data connection.
In general, because of the time and bandwidth required to send the images to the cloud, the processing time may be short, but the data transmission time can be long, depending on the server location and network speed.
Cloud-based versus Edge-based use case
Edge computing / Juven approach
By now you are probably familiar with the Internet of Things. As devices become smarter and smarter, it becomes possible to embed the devices with some degree of AI-vision processing capacity. This allows us to leveraging distributed AI vision in IoT for event management.
That being said, this approach clearly takes more engineering effort for both the hardware, software and their integration. This increases the hardware costs incurred but drastically lowers manual cost and is of a higher efficiency.
How it works in broad strokes
Step 1: Put together the following hardware items: industry-grade camera, a processing unit (aka a computer), an AI chip (to run algorithms offline), a screen (for system control and user interaction), a router (to connect to the Internet), and the holding kiosk/stand.
Step 2: Write into the AI chip the algorithms for facial recognition
Step 3: Integrate the AI chip with the camera and the processing unit
Step 4: Write an application that can run on the processing unit
Step 5: Run the hardware set and detect faces real-time without any admin operations.
Step 6: The facial recognition process takes place within the hardware not on cloud. No data needs to be sent to cloud during the recognition process. The result is shown on the screen directly and immediately, a non-image, usually a binary response (small in data size) is sent back to the cloud-based software (e.g. Juven’s engagement platform) for data integration or further processing.
These new developments are very exciting, and Juven is proud to be working on providing a state-of the art AI-vision enabled check-in process that is available not just to large corporation, but also to consumers as well. Stay tuned for future posts about the topic.
Links:
Juven builds integrated organization software that empowers charities, companies, associations and clubs of all sorts to build, engage and grow community. We believe that organizations should have access to beautiful, intuitive and reliable tools to create greater impact.
